Drone calibration flights
Written by Fabien Vivier
As part of our collaborative project with NOAA PIFSC, JIMAR and Sea Life Park entitled "Calibrating UAS photogrammetry to derive delphinid demographic parameter", we recently collected photographs and videos of objects of known lengths with our Unoccupied Aerial Systems (UAS) for calibration.
Calibration of such instrumentation is done by scaling distances to objects of known sizes by using both flight parameters (altitude, roll, pitch of the drone) and lens distortion parameters. This allows to test precision and accuracy in measuring these objects.
Our system is equipped with:
- A 4/3 gimballed camera mounted with a low-distortion lens (Olympus 25mm, f/1.8)
- An independently powered LIDAR/GPS data-logging system which records altitude and GPS location at 1Hz.
First, we needed to find the perfect flat field for this, and we are thankful for Joshua Levy from the Applied Research Laboratory at the University of Hawaii at Manoa for helping us out with this. Second, we thank the University of Hawaii at Manoa for letting us use their Les Murakami baseball Stadium which was ideal for our purposes.
Following the methodology (*see below for reference details) developed by our colleagues Professor Steve Dawson and Eva Leunissen from the Department of Marine Science in the Biology and Conservation of Marine Mammals and Pascal Sirguey from the School of Surveying at the University of Otago (New Zealand), we have set up a 20 m by 14 m grid made of 88 high-contrast targets. All targets were accurately spaced 2 m of each other using a measuring tape.
Once the grid was set up, we added to it three long poles of known length to help us calibrate our system. We then flew our UAS at altitudes of 20 m, 30 m and 40 m and recorded both photos and videos every 45° in a converging pattern around the grid.
Aerial views of the grid for the UAS calibration.
Our colleagues from the University of Otago in New Zealand will now perform the coding (thanks Eva, Pascal and Steve!!) in order to calibrate the measurements collected by computing/incorporating this new data into their custom-written software.
This calibration process is extremely important and is the first step ofcour research, as inaccurate measurements will lead to inaccurate results. Subsequently, we will complement these calibrations by comparing UAS vs physical measurements of bottlenose dolphins housed at Sea Life Park, Oahu (https://www.mmrphawaii.org/uav-validation) in order to assess precision and accuracy of our instruments in measuring animals. Once we have ensured that the calibration measurements are accurate and that our system is fine-tuned, it will provide confidence in measuring free-ranging cetaceans.
* Reference details: Dawson SM, Bowman MH, Leunissen E and Sirguey P (2017) Inexpensive Aerial Photogrammetry for Studies of Whales and Large Marine Animals. Front. Mar. Sci. 4:366. doi: 10.3389/fmars.2017.00366